Built independently by an author, for readers. Read the story and support ChapterPal
ChapterPal Logo
Log inSign up

keyword

zero-shot document retrieval

Zero-shot document retrieval is an information retrieval approach in which a system searches a corpus to find documents relevant to a user query without having received task-specific or domain-specific training on labeled query-document pairs. Unlike traditional supervised retrieval methods that require fine-tuning on annotated relevance datasets for each target setting, zero-shot retrieval evaluates relevance by leveraging general-purpose pre-trained language models, prompt-guided representations, or unsupervised ranking techniques. This capability allows search systems to generalize directly to unseen domains, diverse tasks, and new document collections without the need for specialized training data.

1 item

PromptReps: Prompting Large Language Models to Generate Dense and Sparse Representations for Zero-Shot Document Retrieval

PromptReps: Prompting Large Language Models to Generate Dense and Sparse Representations for Zero-Shot Document Retrieval

Shengyao Zhuang, Xueguang Ma, Bevan Koopman, Jimmy Lin, Guido Zuccon

OrganizationsCommonwealth Scientific and Industrial Research OrganisationUniversity of QueenslandUniversity of Waterloo

Why you should read this

Introduces a prompt-based approach that simultaneously extracts dense embeddings and sparse bag-of-words representations from large language models in a single forward pass, enabling zero-shot full-corpus document retrieval without expensive contrastive fine-tuning.

Utilizing large language models (LLMs) for zero-shot document ranking is done in one of two ways: (1) prompt-based re-ranking methods, which require no further training but are only feasible for re-ranking a handful of candidate documents due to computational costs; and (2) unsupervised contrastive trained dense retrieval methods, which can retrieve relevant documents from the entire corpus but require a large amount of paired text data for contrastive training. In this paper, we propose PromptReps, which combines the advantages of both categories: no need for training and the ability to retrieve from the whole corpus. Our method only requires prompts to guide an LLM to generate query and document representations for effective document retrieval. Specifically, we prompt the LLMs to represent a given text using a single word, and then use the last token’s hidden states and the corresponding logits associated with the prediction of the next token to construct a hybrid document retrieval system. The retrieval system harnesses both dense text embedding and sparse bag-of-words representations given by the LLM. Our experimental evaluation on the MSMARCO, TREC deep learning and BEIR zero-shot document retrieval datasets illustrates that this simple prompt-based LLM retrieval method can achieve a similar or higher retrieval effectiveness than state-of-the-art LLM embedding methods that are trained with large amounts of unsupervised data, especially when using a larger LLM.

Added

2026-09-26